How to average MFD MPG's

My now deceased father taught me to keep a little notebook in the glovebox. Write down your odometer reading & gallons each time you stop for gas. It seems that way before hybrids, people would buy big pickups or sedans. Fill 'em way up on one tank, then stop at the first click on the next fill-up & brag about
what good mileage they were getting...

Sorry if that sound way low tech. But looking at six months or a year is sound.

I have heard talk of the Prius having a bladder. I topped off today. I had almost exactly 500 miles.
The mileage thing-a-ma-jingy said 62.7. I calculate 500/8 = 62.5 so that is close enough for government work.

I am sure I have not helped you, since you sound more computer savvy than me!

 

I use a spreadsheet to keep track of the statistics of every fill-up. To calculate the life-time average MPG as displayed by the MFD insert a formula that calculates it based on a weighted average. As you correctly stated a weighted average is required since mileage driven varies from tank to tank. I use a weighted average based on the mileage given for any given tank and the total lifetime mileage driven.

Once you have the correct formula just copy it down for every new fill-up.

 

For years I've kept a small notebook in the side door pocket of each of my cars, and at every fillup record the date, odo reading, total $ amount & gallons purchased. This all goes into a spreadsheet to keep track of MPG, miles/week and year, and several other factors. I calculate total lifetime MPG, but what I find most useful is a column that shows the average of the last four tanks. I did this to smooth out the error due to every fill-up not bringing the tank to the exact same 'full' level.

(I actually started writing a perl cgi version of the spreadsheet, though I haven't gotten around to finishing that yet.)

 

or you can do it the super easy way and log it online at http://www.greenhybrid.com and it does it for you.

 

<div class='quotetop'>QUOTE(morpheusx @ Aug 23 2007, 11:06 PM) [snapback]501363[/snapback]</div>

Don't you still have to type in the information? <_<

 

<div class='quotetop'>QUOTE(GeoDosch @ Aug 23 2007, 11:14 PM) [snapback]501368[/snapback]</div>

all you have to type is
Miles Driven and MPG from MFD display can't get any easier then that B)

 

<div class='quotetop'>QUOTE(morpheusx @ Aug 24 2007, 08:26 AM) [snapback]501485[/snapback]</div>

If you're entering MPG off the MFD, then you don't really even need miles driven.

The reason I calculate it based on miles and fueling amount is that I know it's correct (within the accuracy of the gas pump meter & odometer, which I can calibrate.) I don't know how accurate the MFD is. And I've used this method on all my cars for years, most of which had no onboard MPG calculation.

With my spreadsheet, I'm entering 4 values, though I'm also getting a lot more information from that, such as how many miles I'm averaging per day/week/year, how much I've been paying, gas price trends, cost per mile etc.

<div class='quotetop'>QUOTE(morpheusx @ Aug 24 2007, 08:26 AM) [snapback]501485[/snapback]</div>

Sure it could. There could be a link via my bluetooth data connection to automatically upload my mileage data. :lol:

 

I didn't do a spreadsheet for my first year or so but recently started one just to test the accuracy of the MFD. I am doing it the old fashioned way by simply calculating each tank by hand and comparing my numbers to the MFD and come up with +/- average over time. I try to make a point to keep it as consistent as possible by filling up at 2 pips which is usually around 8 gallons. In the short time I have been doing this I have found that the MFD is generous but it varies so much tank to tank I expect that it might even out eventually. Given the argument presented by SureValla I might have to rethink this as an accurate method to test. I never thought I would take advice from a Yankee fan.

 

<div class='quotetop'>QUOTE(SureValla @ Aug 23 2007, 06:40 PM) [snapback]501263[/snapback]</div>

This number really doesn't tell you anything about what you can expect from your next tank, and the reason it doesn't give you the lifetime mpg really doesn't have much to do with the weight of the fuel in the tank. MPG is a ratio that tells you the relationship between miles driven and gallons used, but it doesn't tell you anything about how many miles you drove or how many gallons you used, so simply averaging the ratios only works if they are all computed relative to the same amount.

To calculate lifetime MPG when you regularly reset the MFD you need to keep track of any of the following:
MFD MPG at reset & gallons used for each MPG recorded (MPG * gallons = miles)
MFD MPG at reset & miles driven for each MPG recorded (miles / MPG = gallons)
total number of miles driven (odometer) & total gallons used (total miles / total gallons = lifetime MPG)
<div class='quotetop'>QUOTE(SureValla @ Aug 23 2007, 06:40 PM) [snapback]501263[/snapback]</div>

It really doesn't need to be this complicated.

If you are keeping track of both the MPG and the number of gallons, then the math gets very simple.

MPG * gallons = miles
So for instance 44.2 MPG * 9.93 gallons = 438.906 miles driven on that tank.

Compute miles for each tank, add up all the miles, add up all the gallons, divide the two totals, and you have your lifetime MPG.

44.2 * 9.93 = 438.906 miles
46.2 * 9.95 = 459.69 miles
49.4 * 8.843 = 436.8442 miles

Total miles driven = 1335.4402
Total gallons used = 28.723

1335.4402 mi / 28.723 gal = 46.49376 MPG.

 

<div class='quotetop'>QUOTE(GeoDosch @ Aug 24 2007, 09:43 AM) [snapback]501504[/snapback]</div>

it still needs to know how far you drove with the current average. The miles is a significant part on calculating the lifetime avg
click my green hybrid link below and take a look tank by tank

 

Danny makes a very good point. The prior tank has no relation to the next tank unless your driving conditions are almost identical.

If things change (e.g. temps drop, or you have to climb some hills tomorrow), then the upcoming tank certainly won't match any average you've got.

The same goes for traffic accidents. You can be driving for 20 years without an accident, and that factor often has little to do with what happens in the next mile.

 

<div class='quotetop'>QUOTE(GeoDosch @ Aug 24 2007, 06:43 AM) [snapback]501504[/snapback]</div>

Actually, I enter the miles driven and the gallons pumped, and it figures mine that way, just like by hand!
GreenHybrid allows for different ways of putting in your information.

 

I also thought that the MPG shown is an average for the tank. So you would be averaging an average which would not be correct. I did what was mentioned before and keep a little note book and just record the Date, Odom, (which give me the miles driven), and gas put in to get MPG. Then the anal-yst in me kicked in and I made the spreadsheet (link in my signature). It is a compilation of a few others. I try to update it regularly if you see any mistakes. I have not had to motivation to add any macros yet.

I had thought about doing some correlation testing with the temp data, but I figured I would also need information about city or highway driving, who was driving, etc.

 

Below is a screen-shot of the excel sheet I use. The first four columns are what I enter (date, odo, $$ and gal) and the rest are calculated. The average columns (vs. tank) are for the last four entries, which is to smooth-out variations in fill-ups (i.e. not always refilling to the exact same level.) The gauge column is also calculated, so I can compare it to what my guess gauge was showing (which on mine is actually quite consistent)

[attachmentid=10885]

I've made no attempt to track the type of driving, since over the course of a tank it pretty much runs the gamut, unless I've gone on a long-distance trip, in which case I'll add a comment to the entry.

If anyone is interested I'm happy to share. I also have a variation that I used on my leased vehicles that tracks how far over/under the alloted mileage you are (I'd usually turn in my leased cars just a few miles under the limit.) My online version isn't quite ready for public display.

I've done this for most of my cars (my first version of this was on VisiCalc). It may seem overly involved to some, though as a programmer who grew up with a scientist for a father, it's just the tip of the iceberg.

 

Better than what my wife does.
She records petrol purchased and kilometre readings each fill up but does exactly nothing with the data. Never calculates economy, nothing! She throws the book away when she sells the car.

 

I keep a detailed spreadsheet and the graph files are at Paul's Prius Page. Screenshot below:



I've discovered that accurately averaging my MFD mileage (mutiply miles driven by MPG then divide by total miles) and comparing to total miles divided by total gas purchased shows my MFD reads 1.5 miles better than reality. The MFD is more consistent than the fill-up calc (thanks bladder), but there is a bias in the number.

With that data I can graph things such as this:


or this:

 

<div class='quotetop'>QUOTE(enerjazz @ Aug 25 2007, 12:56 PM) [snapback]502017[/snapback]</div>

That's great how you graph MPG as a function of temperature! I used to do that for my home heating/cooling consumption, but hadn't thought of doing so for my cars. And since the Prius FC is probably affected far more by temperature due to the ICE warm-up phase, it makes perfect sense.

 

I knew I was a bit of a geek, but gosh what a bunch of NERDS!!!!!!!!!!!!!!!


NERDS! NERDS! NERDS! NERDS! NERDS! NERDS! NERDS! NERDS!
NERDS! NERDS! NERDS! NERDS! NERDS! NERDS! NERDS! NERDS!

Yes I just reduced this down to name calling.